Data Quality, 50023

Data Quality, 50023

Nov 11, 2009 by in Data Quality

I am standing at the cash register at the end of the lunch counter at the Peach Pit.  My bill totals $19.90 (including a 10% gratuity).  I swipe my credit card and the tiny LCD screen asks me to enter my billing zip code in order to complete the transaction. 

I slowly enter 5-0-0-2-3 and I am reprimanded by the message: “Incorrect zip code, purchase denied.”  I shake my head and mumble something thankfully incomprehensible to the extremely attractive cashier. 

I then swipe my card a second time and when prompted, I slowly enter 5-0-0-2-1 and I am rewarded by the message: “Purchase approved, have a nice day.”

I take my receipt, slip on my sunglasses, and slowly stroll out of the diner and into the warm California sun where Cindy, Brandon, and Brenda are waiting for me . . .

I don’t live in Beverly Hills, 90210.  I grew up in Everett, 02149 – a northern suburb of Boston, Massachusetts.  I now live in Ankeny – a northern “suburb” of Des Moines, Iowa.  Until a few years ago, the sprawling metropolis that Ankeny isn’t (by any stretch of my overactive imagination) had but a single zip code – 50021. 

The housing boom coupled with a change in zoning laws lead to farm land being purchased and transformed into residential properties.  The United States Postal Service then deemed Ankeny worthy of being divided into two postal zones and me and my fellow citizens of the swanky southern part of the town now found ourselves sporting the new zip code of 50023.

I immediately updated all of my billing addresses, including with the credit card company that guest starred in the reality-based fictional opening scene of this post.

Since I am an obsessive-compulsive data quality expert, you can probably imagine how much it irks me to be informed by tiny LCD screens that I am not providing accurate billing information. 

Since I know quite well that updates to postal certification databases lag behind when the changes actually occur and that not every company applies those updates as soon as they do become available, I was willing to let the issue slide for a little while.

But years have now literally passed since this zip code change has taken place.  And numerous telephone calls to the customer service department of the credit card company in question have provided me with countless assurances that my billing address has been updated in all of their systems.

However, every time I swipe my credit card at the Machine Shed (not quite as nice as the fictional Peach Pit, but still one of my favorite Iowa restaurants none the less), I play data quality roulette with my billing zip code – and I haven’t won a game yet.

I am always interested to hear stories about DQ-IRL (data quality in real life). 

Please share some of your stories. 

Who knows, maybe we can even build a crowdsourced screenplay for a new hit television series.  I think “Data Quality, 50023” sounds like a catchy title . . .

Tags: ,

11 Responses to “Data Quality, 50023”

  1. Henrik Liliendahl Sørensen

    Nov 11, 2009

    Oh yes, having the national character ø in my name (including on my credit cards) often leads to prompts like this on international websites: Please enter a correct name.

    A classic webform error is that you are prompted for a US state. Most often I select Alaska. Guess US business’s have a lot customers up there.

    A have yet to see a correct formatted and spelled address on letters I receive from abroad – and that is those who actually get delivered. By the way: I mostly get overseas letters regarding data quality conferences from organizers and sponsors.

    Reply to this comment

  2. Phil Simon

    Nov 11, 2009

    At Carnegie Mellon during my freshman year, my friend had lost his meal card. A few days later, he attempted to purchase a burger and fries with his replacement card. The message on the register read (and you can’t make this up) “Dead Patron.”

    I’m pretty sure that he’s alive now. He certainly was then.

    Reply to this comment

  3. Dylan Jones

    Nov 11, 2009

    Great post Jim, I had the same problem when I lived on a new development in the UK, the refresh lag on these services can be a total pain and forces you to pass on inaccurate data as you say.

    There is another way to look at this though, through crowd-sourcing innovation.

    Think of all the people in your suburb who are entering those “defective” zips. The errors must be registered by the credit card company somewhere on a database.

    If they were smarter, they could profile repeating occurrences of codes and cross-check their approval either directly with the registration body or simply using a low-cost outsourcer.

    There is an obvious financial benefit here.

    Every time your card is refused it opens up the possibility of a lost sale. Sure you could go through the whole process again but you’re in a rush, you’ve got $10 change in your pocket, screw it I’ll pay by cash this time.

    Credit card company loses 1.5% of the transaction.

    Just imagine how many times a day that must happen. By mining the refusal data for clusters of codes that are occurring far too frequently you’re creating candidates for fast-track validation and approval.

    Same with Henrik’s comment. Are companies trapping the customer details that fail? Are they mining this information to see why data is being abused on entry? Are they adapting their forms accordingly?

    The data never lies and will tell them how to design web forms correctly.

    And if that doesn’t work they should just read Graham Rhind’s excellent ebook: “Better data quality from your web form” http://bit.ly/2zQrPZ

    Great post Jim.

    Reply to this comment

  4. Jim Harris

    Nov 11, 2009

    Thanks for sharing Henrik,

    As someone barely literate in American English, I have always wondered about that symbol – it seemed as if you were attempting to strike the letter O from your name ;-)

    Yes, web forms are great sources of DQ-IRL and Graham Rhind shared a story (and recommendations for improvement) in his recent blog post:

    http://grcdi.blogspot.com/2009/10/dear-reader-youll-know-by-now-that-best.html

    Best Regards,

    Jim

    Reply to this comment

  5. Graham Rhind

    Nov 11, 2009

    Data quality in real life? Never mind a series, clear the schedules for a daily soap opera!

    I’ve been eagerly (25 years) waiting for a foreign company (especially those claiming to be data quality companies) to get my address format correct, and, like Hendrik, each one received is an episode in itself. But recently I received a communication with a correctly formatted address from the US – an absolute primeur, so a big thank you and thumbs up to Address Vision (http://www.addressvisioninc.com/).

    I hope it’s the start of a trend …. :-)

    Reply to this comment

  6. Crysta Anderson

    Nov 11, 2009

    I once lived in an apartment that didn’t exist, according to the US Postal database. My landlord had carved out the attic into a very nice one bedroom apartment, but never bothered to update the post office of its existence. Instead, he just added another mailbox to the front door and slapped a #4 on it. Most mail got delivered just fine. But I usually had to have online orders shipped to my Initiate address since any system that checked against the official USPS database returned an error.

    Now I live in a house that someone had years ago illegally converted into a two-unit. (All physical evidence of that second unit is long gone.) I still occasionally get mail addressed to phantom residents in Apt #2, and the phone company never fails to provide a stack of those big yellow paperweights.

    Reply to this comment

  7. Jim Harris

    Nov 11, 2009

    Thanks everyone for contributing your stories!

    Not to make this a competition or anything, but I have to give the edge to the stories from Phil and Crysta as my favorites so far.

    Anyone else want to share a story about DQ-IRL?

    Best Regards,

    Jim

    Reply to this comment

  8. Charles Blyth

    Nov 11, 2009

    Back in the ‘old days’ in South Africa all school-leaving males had to do military service. I have a brother who is two years older than me. He went to Uni straight from school, and was still there when I went into the military. Much to my delight that when I reported for service, they had me under my brothers name and thought I was two years older than I was. Now, the good thing about that was that I was sent into a unit of men who were fresh from Uni, and were all earmarked for officers school! Talk about a bonus, I had a foot in the door to becoming an officer, while my mates from school were all lumped in with the rest of the riff-raff. And I ended up commanding some of them :) Sometimes DQ-IRL issues can be a good thing.

    PS: They never did correct it, at least they had my name right, it was shouted at me often enough …

    Reply to this comment

  9. Graham Rhind

    Nov 11, 2009

    Well, as United States’ social security numbers were mentioned in today’s IAIDQ Ask The Expert panel presentation as a way to identify individuals, how about this one:

    When I moved to The Netherlands I was assigned a SOFI number, the equivalent to a US social security number. I later moved to Belgium and then, via a circuitous route, ended up again in The Netherlands. At that point some program or person made an error and assigned a new SOFI number to me instead of giving back the old one.

    The authorities have now lost all track of me in the 5 years I was in this country between 1984 and 1989 because their systems can only search for data on the basis of a SOFI number nobody knows any more, with the consequence that I could lose 5 year’s worth of pension payments. A major example of broken processes, and proof that data quality is not a business issue – it affects real people.

    Reply to this comment

  10. Marianne Colwell

    Nov 11, 2009

    I recently purchased a copy of a well known Body of Knowledge (I’ll leave the culprit organisation un-named).

    The purchase was for an electronic PDF, and licensing is controlled by adding the watermark “Licensed by ” to the header of every page. On the web site, I entered my own name, address, email etc. then proceeded to the payment page. One of the payment methods was PayPal. Being lazy, I didn’t want to dig out my credit card, so I elected to use Paypal, which is actually set up under my husband’s name.

    A few minutes later the PDF arrived in my inbox. Aren’t the wonders of technology great? Only problem was: when I opened the PDF, my husband’s name was proudly displayed as the licensee at the header of every page.

    As someone who works closely with customer data in the IP Rights world, I wonder what our customers would think if we issued their IP Right in the name of the payer, rather than the owner?

    Reply to this comment

  11. Jim Harris

    Nov 12, 2009

    Thanks Charles, Graham, and Marianne,

    Your DQ-IRL stories are great!

    I think Graham was right in his earlier comment that “data quality in real life” deserves more than just one series. We should book the schedule of an entire major television network!

    However, we might have to settle for me running a recurring DQ-IRL blog post series on this community.

    Please continue sharing.

    Thanks and Best Regards,

    Jim

    Reply to this comment

Leave a Reply